China’s robot dog helps blind people with 90 percent speech recognition accuracy

A team of researchers in China has developed a robot dog to help visually impaired people navigate.

The six-legged robot, developed by the School of Mechanical Engineering at Shanghai Jiaotong University, uses cameras and sensors to operate and recognize traffic signals.

The robot’s current speech recognition accuracy is over 90 percent and its response time is less than one second. This enables it to quickly and correctly respond to and understand voice commands from visually impaired people.

According to the researchers, clever two-way communication allows the robot to simultaneously issue voice commands and provide real-time feedback on its environment and gait.

Affordable guide dog technology

In China, there are just over 400 guide dogs for nearly 20 million blind people. Keeping pets and service animals is a relatively new concept in the country, so many workplaces, restaurants and other public spaces are not yet welcoming to traditional helpers such as Labradors.

According to the researchers, unlike these dogs, whose supply is limited due to natural breeding constraints and the intensive training required, the production of robotic guide dogs could be achieved on a large scale, especially in a major manufacturing center such as China.

“It’s a bit like cars. I can mass produce them in the same way as cars, so they become more affordable. I think this could be a very big market because there could be millions of people in the world who need guide dogs,” Professor Gao Feng, head of the research team at the institute’s mechanical engineering department, told Reuters.

According to experts, the use of real guide dogs is limited mainly due to the high costs of around $50,000 and the extensive training required, which lasts two to three years.

Furthermore, only about half of the dogs trained will actually help visually impaired people. Guide dogs have the potential to significantly reduce costs, increase efficiency and improve accessibility for those in need.

Robots with improved accessibility

The team’s robot dog is about the size of an English bulldog but slightly wider than a real dog. It has six legs instead of four, which researchers say improves its stability and results in smoother movements.

The robot reaches a maximum speed of 3 meters per second and is suitable for all needs, from slow walking to running. Its unique six-legged design ensures stable, quiet movements.

Currently, the guidance robot is in the field testing phase and is being evaluated through offline demonstrations and functional tests with visually impaired participants. The development team plans to refine and improve the robot based on real-time feedback from these users.

According to the researchers, the implementation of the guidance robot goes beyond the device itself; it is based on big data support in the backend, a robust operation and maintenance team, and extensive advertising testing.

In a similar project, engineers from the computer science department at Binghamton University in New York have developed a guide dog to help people with visual impairments gain access. Last year, they tested the four-legged robot in a trick-or-treating exercise.

After nearly a year of development, the team developed a novel leash-pulling interface using reinforcement learning. After about 10 hours of training, these robots can navigate indoors, guide people, avoid obstacles, and detect pulling.

Engineers see great potential in the pulling interface, which allows users to control the robot by pulling it in a specific direction at hallway intersections, but they believe further research and development is needed before the technology can be used effectively in specific contexts.